Skip to contents

Analysis objectives

  1. Import, recode, and subset data from bugsigdb.org
  2. Create a table of studies
  3. Calculate the frequency of appearance of each taxa in independent signatures and identify the most frequently reported taxa
  4. Estimate the probability of the most frequently identified taxa occuring by chance

Making sure packages are installed

install.packages(c("devtools", "tidyverse", "kableExtra"))
devtools::install_github("waldronlab/BugSigDBStats")
devtools::install_github("waldronlab/bugsigdbr")

Load and subset data

## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr     1.2.1     ✔ readr     2.2.0
## ✔ forcats   1.0.1     ✔ stringr   1.6.0
## ✔ ggplot2   4.0.3     ✔ tibble    3.3.1
## ✔ lubridate 1.9.5     ✔ tidyr     1.3.2
## ✔ purrr     1.2.2     
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors
## 
## Attaching package: 'kableExtra'
## 
## The following object is masked from 'package:dplyr':
## 
##     group_rows
library(bugSigSimple)
dat <- bugsigdbr::importBugSigDB(cache= TRUE)
## Using cached version from 2026-11-09 16:32:08
dim(dat)
## [1] 7425   51
names(dat)
##  [1] "BSDB ID"                    "Study"                     
##  [3] "Study design"               "PMID"                      
##  [5] "DOI"                        "URL"                       
##  [7] "Authors list"               "Title"                     
##  [9] "Journal"                    "Year"                      
## [11] "Keywords"                   "Experiment"                
## [13] "Location of subjects"       "Host species"              
## [15] "Body site"                  "UBERON ID"                 
## [17] "Condition"                  "EFO ID"                    
## [19] "Group 0 name"               "Group 1 name"              
## [21] "Group 1 definition"         "Group 0 sample size"       
## [23] "Group 1 sample size"        "Antibiotics exclusion"     
## [25] "Sequencing type"            "16S variable region"       
## [27] "Sequencing platform"        "Data transformation"       
## [29] "Statistical test"           "Significance threshold"    
## [31] "MHT correction"             "LDA Score above"           
## [33] "Matched on"                 "Confounders controlled for"
## [35] "Pielou"                     "Shannon"                   
## [37] "Chao1"                      "Simpson"                   
## [39] "Inverse Simpson"            "Richness"                  
## [41] "Signature page name"        "Source"                    
## [43] "Curated date"               "Curator"                   
## [45] "Revision editor"            "Description"               
## [47] "Abundance in Group 1"       "MetaPhlAn taxon names"     
## [49] "NCBI Taxonomy IDs"          "State"                     
## [51] "Reviewer"

Subsetting

Subsetting only studies done on humans - 6 of 17 studies were excluded because they were either done on mice or not statistically reliable (Cregger et. al)

included.pmid <-
  c(
    "30778155",
    "32192080",
    "31087436",
    "26901400",
    "33839907",
    "32046455",
    "33925708",
    "32299442",
    "33313185",
    "34268384",
    "33660232"
  )
subset.dat <- dat[dat$PMID %in% included.pmid, ]

All studies

Summary of studies and most frequent taxa increased and decreased in endometriosis patients for all studies

bugSigSimple::createStudyTable(subset.dat) %>%
  kbl() %>%
  kable_paper("hover", full_width = FALSE) %>%
  scroll_box(width = "100%", height = "400px")
Study Identifier MaxCases MaxControls Study design Condition N_signatures PMID DOI URL
AkiyamaK._2019 30 39 cross-sectional observational, not case-control Endometriosis 1 31087436 https://doi.org/10.1111/aji.13147 NA
AtaB._2019 14 14 prospective cohort Endometriosis 2 30778155 https://doi.org/10.1038/s41598-019-39700-6 https://pubmed.ncbi.nlm.nih.gov/30778155/
ChaoX._2021 37 66 case-control Endometriosis 4 34268384 https://doi.org/10.21037/atm-20-4586 NA
ChenS._2020 14 120 case-control Adenomyosis,Endometriosis 3 33313185 https://doi.org/10.21037/atm-20-1309 NA
KhanK.N._2016 32 32 case-control Endometriosis 5 26901400 https://doi.org/10.1016/j.ejogrb.2016.01.040 NA
LeeS.R._2021 45 45 case-control Endometriosis 2 33925708 https://doi.org/10.3390/ijms22094608 NA
PerrottaA.R._2020 8 12 cross-sectional observational, not case-control Endometriosis 2 32046455 https://doi.org/10.1007/s43032-019-00113-5 NA
ShanJ._2021 12 12 case-control Endometriosis 2 33839907 https://doi.org/10.1007/s00404-021-06057-z NA
SvenssonA._2021 66 198 case-control Endometriosis 4 33660232 https://doi.org/10.1007/s43032-021-00506-5 NA
WeiW._2020 36 14 cross-sectional observational, not case-control Endometriosis 2 32299442 https://doi.org/10.1186/s12941-020-00356-0 NA
BugSigDBStats::getMostFrequentTaxa(subset.dat, n=30) %>%
  kbl() %>%
  kable_paper("hover", full_width = FALSE) %>%
  scroll_box(height = "300px")
Var1 Freq
1301 6
1578 4
543 4
84107 4
119852 3
1300 3
135621 3
171549 3
1940338 3
28050 3
39948 3
469 3
838 3
906 3
1350 2
1380 2
1508657 2
1582879 2
1654 2
1663 2
168808 2
1743 2
194 2
216851 2
2701 2
286 2
33038 2
33958 2
419014 2
72294 2
BugSigDBStats::getMostFrequentTaxa(subset.dat, direction="UP") %>%
  kbl() %>%
  kable_paper("hover", full_width = FALSE) %>%
  scroll_box(height = "300px")
Var1 Freq
1301 5
543 4
119852 3
1300 3
135621 3
1578 3
1940338 3
469 3
84107 3
1350 2
BugSigDBStats::getMostFrequentTaxa(subset.dat, direction="DOWN") %>%
  kbl() %>%
  kable_paper("hover", full_width = FALSE) %>%
  scroll_box(height = "300px")
Var1 Freq
168808 2
171549 2
28050 2
33958 2
906 2
1185407 1
1301 1
1452070 1
1508657 1
1578 1

Excluding feces samples

Summary of studies and most frequent taxa in only samples from female reproductive tract, excluding feces samples

included.pmid2 <-
  c(
    "30778155",
    "32192080",
    "31087436",
    "26901400",
    "32046455",
    "33925708",
    "32299442",
    "33313185",
    "34268384"
  )
subset.dat2 <- dat[dat$PMID %in% included.pmid2, ]

reproductive_sigs <-
  subset.dat2[which(subset.dat2$`Body site` != "feces" |
                      is.na(subset.dat2$`Body site`)), ]

bugSigSimple::createStudyTable(reproductive_sigs) %>%
  kbl() %>%
  kable_paper("hover", full_width = FALSE) %>%
  scroll_box(width = "100%", height = "400px")
Study Identifier MaxCases MaxControls Study design Condition N_signatures PMID DOI URL
AkiyamaK._2019 30 39 cross-sectional observational, not case-control Endometriosis 1 31087436 https://doi.org/10.1111/aji.13147 NA
AtaB._2019 14 14 prospective cohort Endometriosis 2 30778155 https://doi.org/10.1038/s41598-019-39700-6 https://pubmed.ncbi.nlm.nih.gov/30778155/
ChaoX._2021 37 66 case-control Endometriosis 4 34268384 https://doi.org/10.21037/atm-20-4586 NA
ChenS._2020 14 120 case-control Adenomyosis,Endometriosis 3 33313185 https://doi.org/10.21037/atm-20-1309 NA
KhanK.N._2016 32 32 case-control Endometriosis 5 26901400 https://doi.org/10.1016/j.ejogrb.2016.01.040 NA
LeeS.R._2021 45 45 case-control Endometriosis 2 33925708 https://doi.org/10.3390/ijms22094608 NA
PerrottaA.R._2020 8 12 cross-sectional observational, not case-control Endometriosis 2 32046455 https://doi.org/10.1007/s43032-019-00113-5 NA
WeiW._2020 36 14 cross-sectional observational, not case-control Endometriosis 2 32299442 https://doi.org/10.1186/s12941-020-00356-0 NA
allfreqs <- bugSigSimple::createTaxonTable(reproductive_sigs, n = 20) %>% #could change number
  arrange(I(decreased_signatures - increased_signatures))
incfreqs <- filter(allfreqs, I(increased_signatures - decreased_signatures) > 0)
decfreqs <- filter(allfreqs, I(increased_signatures - decreased_signatures) < 0)
kableExtra::kbl(allfreqs) %>%
  kable_paper("hover", full_width = FALSE)
Taxon Name Taxonomic Level total_signatures increased_signatures decreased_signatures Binomial Test pval
Streptococcus genus 5 4 1 0.37
Coriobacteriaceae family 3 3 0 0.25
Streptococcaceae family 3 3 0 0.25
Enterobacteriaceae family 3 3 0 0.25
Escherichia/Shigella sp. species 3 3 0 0.25
Acinetobacter genus 3 3 0 0.25
Pseudomonadaceae family 3 3 0 0.25
Alloscardovia genus 2 2 0 0.50
Arthrobacter genus 2 2 0 0.50
Atopobium genus 2 2 0 0.50
Staphylococcaceae family 2 2 0 0.50
Faecalibacterium genus 2 2 0 0.50
Ezakiella genus 2 2 0 0.50
Lactobacillus genus 3 2 1 1.00
Dialister genus 3 2 1 1.00
Prevotella genus 3 2 1 1.00
Gardnerella genus 2 1 1 1.00
Propionibacterium genus 2 1 1 1.00
Megasphaera genus 3 1 2 1.00
Lactobacillaceae family 2 0 2 0.50

Load HMP data from curatedMetagenomicData

suppressPackageStartupMessages(library(curatedMetagenomicData))
# hmp_vagina_metadata <- filter(sampleMetadata, study_name == "HMP_2012" & body_site == "vagina")
all_healthy_adult_vagina_metadata <- filter(sampleMetadata, body_site == "vagina" & disease == "healthy" & age_category == "adult")
se <-
  curatedMetagenomicData::returnSamples(all_healthy_adult_vagina_metadata, dataType = "relative_abundance", counts = FALSE)

Get matrices of species and genus relative abundance

allranks <- mia::splitByRanks(se)
#species_relab <- t(assay(allranks[["species"]]))
genus_relab <- t(assay(allranks[["genus"]]))
#family_relab <- t(assay(allranks[["family"]]))

Centered log-ratio transformation on genus level relative abundance:

genus_relab_clr <- Hotelling::clr(genus_relab + 1)
df <- dplyr::select(data.frame(genus_relab_clr), c("Streptococcus", "Lactobacillus"))
df$age <- se$age
df$BMI <- se$BMI
fit <- lm(Streptococcus ~ Lactobacillus + age + BMI, data = df)  
summary(fit)
## 
## Call:
## lm(formula = Streptococcus ~ Lactobacillus + age + BMI, data = df)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -0.6728 -0.1875 -0.0264  0.0893  3.6729 
## 
## Coefficients:
##                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)   -0.386413   0.436694  -0.885 0.378987    
## Lactobacillus -0.078095   0.047933  -1.629 0.107344    
## age            0.035738   0.010073   3.548 0.000665 ***
## BMI           -0.009094   0.014295  -0.636 0.526581    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.4857 on 77 degrees of freedom
##   (14 observations deleted due to missingness)
## Multiple R-squared:  0.1799, Adjusted R-squared:  0.1479 
## F-statistic: 5.629 on 3 and 77 DF,  p-value: 0.00153
summary(lm(Streptococcus ~ Lactobacillus, data = df))
## 
## Call:
## lm(formula = Streptococcus ~ Lactobacillus, data = df)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -0.6006 -0.0609 -0.0462 -0.0440  4.1909 
## 
## Coefficients:
##               Estimate Std. Error t value Pr(>|t|)   
## (Intercept)    0.55246    0.20193   2.736  0.00745 **
## Lactobacillus -0.11409    0.04751  -2.401  0.01832 * 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.5642 on 93 degrees of freedom
## Multiple R-squared:  0.05839,    Adjusted R-squared:  0.04826 
## F-statistic: 5.767 on 1 and 93 DF,  p-value: 0.01832
wilcox.test(df$Streptococcus, df$Lactobacillus)
## 
##  Wilcoxon rank sum test with continuity correction
## 
## data:  df$Streptococcus and df$Lactobacillus
## W = 213.5, p-value < 2.2e-16
## alternative hypothesis: true location shift is not equal to 0
t.test(df$Streptococcus, df$Lactobacillus)
## 
##  Welch Two Sample t-test
## 
## data:  df$Streptococcus and df$Lactobacillus
## t = -28.666, df = 133.93, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -4.258800 -3.709054
## sample estimates:
## mean of x mean of y 
## 0.0879008 4.0718280
secount <-
  curatedMetagenomicData::returnSamples(all_healthy_adult_vagina_metadata, dataType = "relative_abundance", counts = TRUE)
## 
## $`2021-03-31.FerrettiP_2018.relative_abundance`
## dropping rows without rowTree matches:
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Coriobacteriaceae|g__Collinsella|s__Collinsella_stercoris
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Coriobacteriaceae|g__Enorma|s__[Collinsella]_massiliensis
##   k__Bacteria|p__Firmicutes|c__Bacilli|o__Bacillales|f__Bacillales_unclassified|g__Gemella|s__Gemella_bergeri
##   k__Bacteria|p__Firmicutes|c__Bacilli|o__Lactobacillales|f__Carnobacteriaceae|g__Granulicatella|s__Granulicatella_elegans
##   k__Bacteria|p__Firmicutes|c__Clostridia|o__Clostridiales|f__Ruminococcaceae|g__Ruminococcus|s__Ruminococcus_champanellensis
##   k__Bacteria|p__Firmicutes|c__Erysipelotrichia|o__Erysipelotrichales|f__Erysipelotrichaceae|g__Bulleidia|s__Bulleidia_extructa
##   k__Bacteria|p__Proteobacteria|c__Betaproteobacteria|o__Burkholderiales|f__Sutterellaceae|g__Sutterella|s__Sutterella_parvirubra
##   k__Bacteria|p__Spirochaetes|c__Spirochaetia|o__Leptospirales|f__Leptospiraceae|g__Leptospira|s__Leptospira_hartskeerlii
##   k__Bacteria|p__Synergistetes|c__Synergistia|o__Synergistales|f__Synergistaceae|g__Cloacibacillus|s__Cloacibacillus_evryensis
## $`2021-03-31.HMP_2012.relative_abundance`
## dropping rows without rowTree matches:
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Atopobiaceae|g__Olsenella|s__Olsenella_profusa
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Coriobacteriaceae|g__Collinsella|s__Collinsella_stercoris
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Coriobacteriaceae|g__Enorma|s__[Collinsella]_massiliensis
##   k__Bacteria|p__Firmicutes|c__Bacilli|o__Bacillales|f__Bacillales_unclassified|g__Gemella|s__Gemella_bergeri
##   k__Bacteria|p__Firmicutes|c__Bacilli|o__Lactobacillales|f__Carnobacteriaceae|g__Granulicatella|s__Granulicatella_elegans
##   k__Bacteria|p__Firmicutes|c__Clostridia|o__Clostridiales|f__Ruminococcaceae|g__Ruminococcus|s__Ruminococcus_champanellensis
##   k__Bacteria|p__Firmicutes|c__Erysipelotrichia|o__Erysipelotrichales|f__Erysipelotrichaceae|g__Bulleidia|s__Bulleidia_extructa
##   k__Bacteria|p__Proteobacteria|c__Betaproteobacteria|o__Burkholderiales|f__Sutterellaceae|g__Sutterella|s__Sutterella_parvirubra
##   k__Bacteria|p__Synergistetes|c__Synergistia|o__Synergistales|f__Synergistaceae|g__Cloacibacillus|s__Cloacibacillus_evryensis
## $`2021-03-31.WampachL_2018.relative_abundance`
## dropping rows without rowTree matches:
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Coriobacteriaceae|g__Collinsella|s__Collinsella_stercoris
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Coriobacteriaceae|g__Enorma|s__[Collinsella]_massiliensis
##   k__Bacteria|p__Firmicutes|c__Bacilli|o__Lactobacillales|f__Carnobacteriaceae|g__Granulicatella|s__Granulicatella_elegans
##   k__Bacteria|p__Proteobacteria|c__Betaproteobacteria|o__Burkholderiales|f__Sutterellaceae|g__Sutterella|s__Sutterella_parvirubra
##   k__Bacteria|p__Synergistetes|c__Synergistia|o__Synergistales|f__Synergistaceae|g__Cloacibacillus|s__Cloacibacillus_evryensis
allranks <- mia::splitByRanks(secount)
genus_relab_count <- t(assay(allranks[["genus"]]))
df <- dplyr::select(data.frame(genus_relab_count), c("Streptococcus", "Lactobacillus"))
df$age <- secount$age
df$BMI <- secount$BMI
## Classes and Methods for R originally developed in the
## Political Science Computational Laboratory
## Department of Political Science
## Stanford University (2002-2015),
## by and under the direction of Simon Jackman.
## hurdle and zeroinfl functions by Achim Zeileis.
fit.ZInegbin <- pscl::zeroinfl(Streptococcus ~ Lactobacillus + age + BMI | 1,
                data = df,
                dist = "negbin")

Create a data frame for regression analysis from healthy controls for log-linear model

library(curatedMetagenomicData)
library(dplyr)
library(DT)
library(purrr)

taxon_strep <- c("1884" = "Streptococcus")
taxon_lact <- c("1578" = "Lactobacillus")

target_sample <- "vagina"
sample_metadata <- sampleMetadata %>% 
    filter(
        grepl(target_sample, body_site) | grepl(target_sample, body_subsite)
    ) %>% 
    discard(~all(is.na(.x))) %>% 
    as_tibble()
dim(sample_metadata)
## [1] 96 26
unique(sample_metadata$study_name)
## [1] "FerrettiP_2018" "HMP_2012"       "WampachL_2018"
tse <- returnSamples(
    sampleMetadata = sample_metadata,
    dataType = "relative_abundance",
    counts = FALSE,
    rownames = "short"
)
## 
## $`2021-03-31.FerrettiP_2018.relative_abundance`
## dropping rows without rowTree matches:
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Coriobacteriaceae|g__Collinsella|s__Collinsella_stercoris
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Coriobacteriaceae|g__Enorma|s__[Collinsella]_massiliensis
##   k__Bacteria|p__Firmicutes|c__Bacilli|o__Bacillales|f__Bacillales_unclassified|g__Gemella|s__Gemella_bergeri
##   k__Bacteria|p__Firmicutes|c__Bacilli|o__Lactobacillales|f__Carnobacteriaceae|g__Granulicatella|s__Granulicatella_elegans
##   k__Bacteria|p__Firmicutes|c__Clostridia|o__Clostridiales|f__Ruminococcaceae|g__Ruminococcus|s__Ruminococcus_champanellensis
##   k__Bacteria|p__Firmicutes|c__Erysipelotrichia|o__Erysipelotrichales|f__Erysipelotrichaceae|g__Bulleidia|s__Bulleidia_extructa
##   k__Bacteria|p__Proteobacteria|c__Betaproteobacteria|o__Burkholderiales|f__Sutterellaceae|g__Sutterella|s__Sutterella_parvirubra
##   k__Bacteria|p__Spirochaetes|c__Spirochaetia|o__Leptospirales|f__Leptospiraceae|g__Leptospira|s__Leptospira_hartskeerlii
##   k__Bacteria|p__Synergistetes|c__Synergistia|o__Synergistales|f__Synergistaceae|g__Cloacibacillus|s__Cloacibacillus_evryensis
## Duplicated labels were made unique.
## $`2021-03-31.HMP_2012.relative_abundance`
## dropping rows without rowTree matches:
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Atopobiaceae|g__Olsenella|s__Olsenella_profusa
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Coriobacteriaceae|g__Collinsella|s__Collinsella_stercoris
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Coriobacteriaceae|g__Enorma|s__[Collinsella]_massiliensis
##   k__Bacteria|p__Firmicutes|c__Bacilli|o__Bacillales|f__Bacillales_unclassified|g__Gemella|s__Gemella_bergeri
##   k__Bacteria|p__Firmicutes|c__Bacilli|o__Lactobacillales|f__Carnobacteriaceae|g__Granulicatella|s__Granulicatella_elegans
##   k__Bacteria|p__Firmicutes|c__Clostridia|o__Clostridiales|f__Ruminococcaceae|g__Ruminococcus|s__Ruminococcus_champanellensis
##   k__Bacteria|p__Firmicutes|c__Erysipelotrichia|o__Erysipelotrichales|f__Erysipelotrichaceae|g__Bulleidia|s__Bulleidia_extructa
##   k__Bacteria|p__Proteobacteria|c__Betaproteobacteria|o__Burkholderiales|f__Sutterellaceae|g__Sutterella|s__Sutterella_parvirubra
##   k__Bacteria|p__Synergistetes|c__Synergistia|o__Synergistales|f__Synergistaceae|g__Cloacibacillus|s__Cloacibacillus_evryensis
## Duplicated labels were made unique.
## $`2021-03-31.WampachL_2018.relative_abundance`
## dropping rows without rowTree matches:
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Coriobacteriaceae|g__Collinsella|s__Collinsella_stercoris
##   k__Bacteria|p__Actinobacteria|c__Coriobacteriia|o__Coriobacteriales|f__Coriobacteriaceae|g__Enorma|s__[Collinsella]_massiliensis
##   k__Bacteria|p__Firmicutes|c__Bacilli|o__Lactobacillales|f__Carnobacteriaceae|g__Granulicatella|s__Granulicatella_elegans
##   k__Bacteria|p__Proteobacteria|c__Betaproteobacteria|o__Burkholderiales|f__Sutterellaceae|g__Sutterella|s__Sutterella_parvirubra
##   k__Bacteria|p__Synergistetes|c__Synergistia|o__Synergistales|f__Synergistaceae|g__Cloacibacillus|s__Cloacibacillus_evryensis
## Duplicated labels were made unique.
## Warning in convertNode(tree = tree, node = labN, use.alias = FALSE, message =
## FALSE): Multiple nodes are found to have the same label.
tse
## class: TreeSummarizedExperiment 
## dim: 908 96 
## metadata(0):
## assays(1): relative_abundance
## rownames(908): species:Escherichia coli species:Klebsiella
##   michiganensis ... species:Pediococcus pentosaceus species:Rahnella
##   aquatilis
## rowData names(7): superkingdom phylum ... genus species
## colnames(96): CA_C10001MS2003VA_t0M15 CA_C10003MS2031VA_t0M15 ...
##   MV_C118 MV_C120
## colData names(25): study_name subject_id ... family_role body_subsite
## reducedDimNames(0):
## mainExpName: NULL
## altExpNames(0):
## rowLinks: a LinkDataFrame (908 rows)
## rowTree: 1 phylo tree(s) (10430 leaves)
## colLinks: NULL
## colTree: NULL

Comparing mean to variance of the outcome variable

library(ggplot2)
mean(df$Streptococcus)
## [1] 4049.558
var(df$Streptococcus)
## [1] 509068831
fraczeroes <- nrow(df[df$Streptococcus == 0,])/nrow(df)
fraczeroes
## [1] 0.8736842

#Fit log-linear models

fit.pois <- glm(Streptococcus ~ Lactobacillus + age + BMI,
                data = df,
                family = poisson(link = "log"))

summary (fit.pois)
## 
## Call:
## glm(formula = Streptococcus ~ Lactobacillus + age + BMI, family = poisson(link = "log"), 
##     data = df)
## 
## Coefficients:
##                 Estimate Std. Error z value Pr(>|z|)    
## (Intercept)   -5.313e-01  1.873e-02  -28.36   <2e-16 ***
## Lactobacillus  5.141e-08  9.828e-10   52.32   <2e-16 ***
## age            3.599e-01  4.136e-04  870.25   <2e-16 ***
## BMI           -1.573e-01  6.453e-04 -243.77   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for poisson family taken to be 1)
## 
##     Null deviance: 1929049  on 80  degrees of freedom
## Residual deviance:  632598  on 77  degrees of freedom
##   (14 observations deleted due to missingness)
## AIC: 632704
## 
## Number of Fisher Scoring iterations: 20
library(pscl)
fit.ZIpois <- zeroinfl(Streptococcus ~ Lactobacillus + age + BMI | 1,
                data = df,
                dist = "poisson")

summary (fit.ZIpois)
## 
## Call:
## zeroinfl(formula = Streptococcus ~ Lactobacillus + age + BMI | 1, data = df, 
##     dist = "poisson")
## 
## Pearson residuals:
##     Min      1Q  Median      3Q     Max 
## -0.3753 -0.3753 -0.3752 -0.3750 27.8168 
## 
## Count model coefficients (poisson with log link):
##                 Estimate Std. Error z value Pr(>|z|)
## (Intercept)    6.169e+00        NaN     NaN      NaN
## Lactobacillus  1.321e-07        NaN     NaN      NaN
## age            2.420e-01        NaN     NaN      NaN
## BMI           -2.116e-01        NaN     NaN      NaN
## 
## Zero-inflation model coefficients (binomial with logit link):
##             Estimate Std. Error z value Pr(>|z|)
## (Intercept)     1.96        NaN     NaN      NaN
## 
## Number of iterations in BFGS optimization: 20 
## Log-likelihood: -1.02e+05 on 5 Df
## 
## Attaching package: 'MASS'
## The following object is masked from 'package:dplyr':
## 
##     select
fit.negbin <-
  glm.nb(
    Streptococcus ~ Lactobacillus + age + BMI,
    data = df,
    control = glm.control(
      epsilon = 1e-8,
      maxit = 50,
      trace = FALSE
    )
  )
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning: glm.fit: algorithm did not converge
## Warning in theta.ml(Y, mu, sum(w), w, limit = control$maxit, trace =
## control$trace > : iteration limit reached
## Warning in sqrt(1/i): NaNs produced
## Warning in glm.nb(Streptococcus ~ Lactobacillus + age + BMI, data = df, :
## alternation limit reached
library(broom)
tidy(fit.negbin)
## # A tibble: 4 × 5
##   term               estimate   std.error statistic     p.value
##   <chr>                 <dbl>       <dbl>     <dbl>       <dbl>
## 1 (Intercept)   -34.4         8.11            -4.24 0.0000224  
## 2 Lactobacillus   0.000000889 0.000000196      4.54 0.00000570 
## 3 age             1.09        0.207            5.30 0.000000119
## 4 BMI             0.297       0.277            1.07 0.282
fit.ZInegbin <- zeroinfl(Streptococcus ~ Lactobacillus + age + BMI | 1,
                data = df,
                dist = "negbin")

summary(fit.ZInegbin)
## 
## Call:
## zeroinfl(formula = Streptococcus ~ Lactobacillus + age + BMI | 1, data = df, 
##     dist = "negbin")
## 
## Pearson residuals:
##     Min      1Q  Median      3Q     Max 
## -0.1923 -0.1923 -0.1923 -0.1923  7.1121 
## 
## Count model coefficients (negbin with log link):
##                 Estimate Std. Error z value Pr(>|z|)
## (Intercept)    1.917e+01        NaN     NaN      NaN
## Lactobacillus -2.725e-07        NaN     NaN      NaN
## age            3.274e-02        NaN     NaN      NaN
## BMI           -4.033e-01        NaN     NaN      NaN
## Log(theta)    -9.618e-01        NaN     NaN      NaN
## 
## Zero-inflation model coefficients (binomial with logit link):
##             Estimate Std. Error z value Pr(>|z|)
## (Intercept)     1.91        NaN     NaN      NaN
## 
## Theta = 0.3822 
## Number of iterations in BFGS optimization: 27 
## Log-likelihood: -131.9 on 6 Df